This study assessed a quantitative PCR (qPCR) assay for Pneumocystis jirovecii quantification in bronchoalveolar lavage (BAL) fluid samples combined with serum (1¡3)-␤-D-glucan (BG) level detection to distinguish Pneumocystis pneumonia (PCP) from pulmonary colonization with P. jirovecii. Forty-six patients for whom P. jirovecii was initially detected in BAL fluid samples were retrospectively enrolled. Based on clinical data and results of P. jirovecii detection, 17 and 29 patients were diagnosed with PCP and colonization, respectively. BAL fluid samples were reassayed using a qPCR assay targeting the mitochondrial large subunit rRNA gene. qPCR results and serum BG levels (from a Fungitell kit) were analyzed conjointly. P. jirovecii DNA copy numbers were significantly higher in the PCP group than in the colonization group (1.3 ؋ 10 7 versus 3.4 ؋ 10 3 copies/l, P < 0.05). A lower cutoff value (1.6 ؋ 10 3 copies/l) achieving 100% sensitivity for PCP diagnosis and an upper cutoff value (2 ؋ 10 4 copies/ l) achieving 100% specificity were determined. Applying these two values, 13/17 PCP patients and 19/29 colonized patients were correctly assigned to their patient groups. For the remaining 14 patients with P. jirovecii DNA copy numbers between the cutoff values, PCP and colonization could not be distinguished on the basis of qPCR results. Four of these patients who were initially assigned to the PCP group presented BG levels of >100 pg/ml. The other 10 patients, who were initially assigned to the colonization group, presented BG levels of <100 pg/ml. These results suggest that the combination of the qPCR assay, applying cutoff values of 1.6 ؋ 10 3 and 2 ؋ 10 4 copies/l, and serum BG detection, applying a 100 pg/ml threshold, can differentiate PCP and colonization diagnoses.

The results provide to our knowledge the first data on the role of colonized patients as potential sources of P. jirovecii in a context of nosocomial acquisition of the fungus.